Vegetation and
the terrestrial
carbon cycle:
Modelling the first
400 million years
D. J. B e e r l i n g & F. I . W o o d w a r d
Department of Animal and Plant Sciences, University of Sheffield, Sheffield S10 2TN, U.K.
published by the press syndicate of the university of cambridge
The Pitt Building, Trumpington Street, Cambridge, United Kingdom
cambridge university press
The Edinburgh Building, Cambridge CB2 2RU, UK
40 West 20th Street, New York, ny 10011–4211, USA
10 Stamford Road, Oakleigh, vic 3166, Australia
Ruiz de Alarcón 13, 28014 Madrid, Spain
Dock House, The Waterfront, Cape Town 8001, South Africa
http://www.cambridge.org
© D.J. Beerling and F.I. Woodward 2001
This book is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.
First published 2001
Printed in the United Kingdom at the University Press, Cambridge
Typeface teff Lexicon 10/14 pt
System QuarkXPress®
[se]
A catalogue record for this book is available from the British Library
Library of Congress Cataloguing in Publication data
Beerling, D. J.
Vegetation and the terrestrial carbon cycle : modelling the first 400 million years /
D.J. Beerling & F.I. Woodward.
p. cm.
Includes bibliographical references (p. ).
I` SBN 0 521 80196 6
1. Carbon cycle (Biogeochemistry)–Mathematical models. 2. Plant
ecology–Mathematical models. 3. Paleoclimatology–Mathematical models.
I. Woodward, F.I. II. Title.
QH344. B43 2001 577.144015118–dc21
ISBN 0 521 80196 6 hardback
Contents
Preface vii
Acknowledgements ix
1 Introduction 1
2 Investigating the past from the present 9
3 Climate and terrestrial vegetation 23
4 Climate and terrestrial vegetation of the present 53
5 The late Carboniferous 100
6 The Jurassic 135
7 The Cretaceous 183
8 The Eocene 238
9 The Quaternary 280
10 Climate and terrestrial vegetation in the future 340
11 Endview 353
References 361
Index 395
v
1
Introduction
Overview
This book considers the operation of the terrestrial carbon cycle
across a range of spatial and temporal scales, with special emphasis on the
photosynthetic carbon metabolism of land plants over the last 400 million
years. Chapters 2 and 3 outline some of the fundamental approaches used in
attempting to predict the operation of plants in the past from a knowledge
of present-day processes, possible means of testing these approaches, and
detailed consideration of the nature of the relationship between vegetation,
soils and climate in a contemporary climatic setting. The later chapters consider a number of ‘globally-averaged’ responses of plant processes to changes
in global mean temperature and atmospheric composition throughout
much of the Phanerozoic. In taking this approach, our view has been to
open up investigation and discussion of how predictions of geochemical
models impinge on the processes governing the regulation of carbon gain
and water loss in leaves of plants, following earlier work of this sort
(Beerling, 1994).
Consideration of the global view is initiated in Chapter 4 with a detailed
discussion of global climate simulations using computers and examining
how these simulations can be modified to predict ancient climates. The discussion is then extended to deal with the development of a model representing the various above- and below-ground processes of the terrestrial carbon
cycle. After extensive satisfactory testing of several key model outputs against
a variety of ground truth data, including observations from satellites, the
stage is set to bring these two global-scale approaches together and investigate changes in the operation of the terrestrial biosphere through geological
time.
The following series of chapters (Chapters 5 to 9) consider, at the global
1
2
Introduction
scale, the effects of several computer climate simulations, representing particular averaged ‘slices’ of geological time, on terrestrial ecosystem properties
and function. The selection of a particular interval for detailed global investigation was constrained by the availability of palaeoclimate simulations from
the climate modelling community. In taking this approach it has been
assumed, we hope not unreasonably, that what constitutes an interesting
interval for climate modellers to test and apply their models represents a correspondingly interesting time to investigate the impact of the modelled climates on plants, vegetation and terrestrial ecosystem carbon cycling.
The geological timescale
Studies of a geological nature necessarily, by definition, deal with
intervals of time typically ranging from thousands to millions of years. These
timescales are difficult for the human mind to appreciate yet are those
required for considering the gradual processes operating to shape the Earth’s
climate and the course and pattern of evolution. Indeed, it is interesting to
note that amongst other political reasons, the very difficulty humans have of
comprehending and accepting a time span of thousands to millions of years
bedevilled our ability to formulate the concept of so-called ‘deep time’ in the
first place. The first realisation that the Earth was in fact millions of years old
was made by James Hutton in his Theory of the Earth (1780s) and was subsequently refined and developed further by Charles Lyell in his Principles of
Geology (1830–1833). With this advancement in hand, Charles Darwin was
able to proceed with the development of his theory of evolution by the accumulation of small gradual changes, via natural selection, over millions of
years.
A basic appreciation of the geological time span and the names of the key
geological eras, periods and epochs is required for this book. For simplicity, we
have refrained from adopting detailed geological names and stratigraphic
nomenclature and used only those more widely agreed upon (Table 1.1). Table
1.1 provides diagrammatic representation of the geological column, note that
it is not scaled equidistantly through time. The Phanerozoic era extends from
544 Ma (million years ago). Its first three periods, the Cambrian, Ordovician
and Silurian encompass the time before the evolution of vascular land plants,
the focus of this text, and have been omitted from Table 1.1. Several global
simulations deal with the Mesozoic era (‘middle life’ or the Age of Dinosaurs),
which encompasses the Triassic, Jurassic and Cretaceous periods. The name
Cretaceous derives from the latin ‘creta’ meaning chalk (widely deposited in
shallow seas during the later stages of the Mesozoic). Following the Mesozoic
The geological timescale
Table 1.1. The geological timetable and key names used in this text
Time in
millions
of years
ago (Ma)
Eon
Era
Period
Quaternary
⇑0
First grasses, first horses, rapid diversification of
mammals.
Cretaceous
Origin and diversification of angiosperms. Extinction of
dinosaurs at the Cretaceous–Tertiary boundary.
Jurassic
Mesozoic
Phanerozoic
Rise of modern ferns, diversification of conifers.
205
Triassic
⇓ to 416
Source: After Haq et al. (1998).
Devonian
354
Carboniferous
Palaeozoic
295
Permian
248
Establishment of regional differences in floras, including
temperate floras and grasslands.
Palaeogene
65
Ice ages.
Neogene
24
Tertiary
Cenozoic
1.8
144
Significant biotic events
Cycads and Ginkgos dominant. Massive biotic extinction
at the Permian–Triassic boundary.
Spread of upland floras into lowlands, demise of lowland
swamps.
Major interval of coal formations, forest dominance by
arborescent lycopods. First land winged insects, first land
vertebrates.
First seed plants, first sharks. Diversification of plants.
3
4
Introduction
is the Cenozoic era (‘recent life’ or the Age of Mammals) which includes the
Tertiary, a term remnant from when Earth history was divided into four intervals: primary, secondary, tertiary and quaternary. The first two are now out of
use whilst the name Tertiary is used to define the time between the end of the
Cretaceous and the beginning of the most recent succession of ice ages, the
Quaternary. The boundary between the Cretaceous and the Tertiary is termed
the K/T boundary, K from the German word for chalk (‘Kriede’) and T to indicate the Tertiary.
Chapters 8 and 9 deal with changes in biospheric operation during intervals represented by subdivisions of the Tertiary and Quaternary periods,
respectively, and since these require some familiarity with the timing
and names of the different periods and epochs within, these are given in Table
1.2.
The global-scale studies in Chapters 5 to 10 recognise the Earth operating
in, broadly, either an ice age or a greenhouse mode.
Ice age Earth
The Earth has experienced four major phases of glaciation over the
last billion years (Crowley & North, 1991): the late Precambrian (800–600 Ma),
Late Ordovician (440 Ma), Permo-Carboniferous (330–275 Ma) and the late
Cenozoic (40–0 Ma). We consider the impacts of two of these ice age episodes
on the terrestrial carbon cycle at times of widely differing continental configurations: the late Carboniferous (300 Ma) (Chapter 5), and during the last
glacial maximum (LGM) at 18 000 years (18 ka) (Chapter 9). During the
Carboniferous several of the continents sutured to form three major land
masses (Gondwana, Eurasia and Kazakhstan). These were configured in very
different locations from the present day, with the land masses cutting across
many parts of zonal circulation, a large portion of land exposed in the low
mid-latitudes, and the presence of a warm seaway (Tethys). All of these features combined to exert major effects on the operation of the global climate
system with a resultant marked difference between Palaeozoic and contemporary climates (Parrish, 1993).
A further and important difference between the late Quaternary ice ages
and those of the Permo-Carboniferous lies in the atmospheric composition.
Late Quaternary ice core records indicate that the LGM was characterised by
low concentrations of atmospheric CO™ (180–200 ppm); and geochemical
modelling of the long-term carbon cycle, as well as stomatal-based palaeoCO™ estimates, suggest a low value for the Carboniferous (300 ppm) (reviewed
in Berner, 1998). So both glaciations are likely to have resulted in part from
Ice age Earth
Table 1.2. Summarised geological timetable of the Cenozoic era
Time in
millions
of years
ago (Ma)
Era
Period
The last 10 000 years before present (ka BP),
including a climatic optimum around 6 ka BP.
Pliestocene
Quaternary
0.01
Climatic comments
Holocene
⇑0.0
Epoch
Interval of repeated glacial–interglacial cycles
1.8
Pliocene
Closure of Panamanian seaway, with profound
changes in deep Atlantic circulation at c. 4.6 Ma.
4.9
Miocene
Oligocene
Tertiary
Cenozoic
24.0
Interval of global warmth without evidence of high
atmospheric CO™ content. Global expansion of
grasslands with the C¢ photosynthetic pathway in
the late Miocene.
Episodes of large fluctuations in global sea level
and polar ice sheet growth and retreat.
34.0
Eocene
Onset of the formation of Antarctic bottom water.
India collides with Asia.
54.0
Palaeocene
⇓65.0
Abrupt climatic warming at the Palaeocene–Eocene
boundary.
Source: After Haq et al. (1998).
the same mechanism, a lowered greenhouse effect. What differs between
them, however, are the processes involved in sequestering carbon from the
atmospheric reservoir and the subsequent effects on the atmospheric O™
content. In the Carboniferous, the massive organic carbon burial, as witnessed by the vast coal swamps of the Carboniferous and early Permian
(330–260 Ma), removed CO™ from the atmosphere, and was calculated to have
5
6
Introduction
been accompanied by a rise in atmospheric O™ content (Berner & Canfield,
1989). During the Quaternary ice ages, CO™ removal from the atmosphere
was mainly driven through changes in Milankovitch orbital insolation lowering global temperatures (Hays et al., 1976) thereby increasing CO™ solubility
into the oceans, but with little effect on their O™ content since oxygen is relatively insoluble in water. Milankovitch insolation variations probably also
operated in the Carboniferous, as revealed by high frequency cyclical sea level
variations in sediments of that age (Wanless & Shepard, 1936), but the effects
of organic carbon burial dominated atmospheric CO™ changes at that time.
Indeed, it is possible to reasonably predict ice sheet location and growth in
the Carboniferous using the same model as that capable of predicting ice
sheet occurrence during Quaternary glacial stages (Hyde et al., 1999). The
low ice-age atmospheric CO™ concentrations place severe constraints on the
efficiency of the photosynthetic metabolism of plants, constraints further
exacerbated by a high Carboniferous O™ content. The underlying reasons for
these constraints in the different periods are discussed in detail in Chapters 5
and 9.
Greenhouse Earth
Three chapters investigate the impact of the Earth’s climate in periods
of what might broadly be described as a greenhouse mode: the Jurassic
(Chapter 6), Cretaceous (Chapter 7) and Eocene (Chapter 8). Chapter 10 deals
with vegetation-climate interactions in a future greenhouse Earth, driven by
CO™ accumulation in the atmosphere following the anthropogenic activities
of fossil fuel burning and deforestation.
There has been a long-standing interest in assessing the major determinants regulating global climate in the distant past, especially at times when
sedimentary rocks and fossilised biota indicate much warmer climates than
now. General circulation models of the Earth’s climate provide a means of
exploring this sensitivity and separating out the relative importance of continental configuration, oceanic circulation and greenhouse gases (e.g. Barron et
al., 1995). There has also been the motivation that simulations of these ancient
greenhouse climates together with similar models used to make predictions
of future global change, offer a means of testing the underlying physics and
other mathematical representations of the Earth’s climate system. However, it
is doubtful (Chapter 8) that any past greenhouse climate provides a realistic
representation for a future greenhouse world.
By analogy, these warm greenhouse climates offer us a means of examining
the productivity of the terrestrial biosphere under such conditions, and
Conclusions
Chapter 8 gives an explicit comparison between the impacts of the most
recent of the ‘ancient’ greenhouse climate episodes, of the Eocene (50 Ma), and
those modelled for the future. Limitations in our knowledge of how the processes represented within the terrestrial carbon cycle model described in
Chapter 4 acclimate, or become modified by long-term exposure to a warm
climate and/or a high CO™ environment, have to be conceded in making these
model predictions. However, meta-analyses of the responses of photosynthesis to elevated CO™ generally indicate that a large and continuing stimulation
of photosynthesis is well described by the Farquhar et al. (1980) photosynthesis sub-model (Medlyn et al., 1999; Peterson et al., 1999) and so this core
process might be regarded as somewhat robust. Further discussion of the
issue of acclimation by the photosynthetic apparatus can be found in Chapter
6, in which the first global simulation with a higher-than-present CO™ atmosphere is presented.
Catastrophic climatic change: the end-Cretaceous mass
extinction
Abundant and globally-widespread geochemical, morphological,
marine and terrestrial evidence has increasingly accumulated, supporting the
idea, first proposed by Alvarez et al. (1980), that a large extraterrestrial bolide
hit the Earth at the end of the Cretaceous. This represents a major perturbation of the global climate system with a correspondingly large expected effect
on the functioning of the biosphere. Terrestrial carbon cycle sensitivity
studies, with modified global late Cretaceous climate datasets, provide one
means of assessing such effects. Therefore, Chapter 7 presents global simulations of vegetation properties and carbon cycling near the end of the
Cretaceous (66 Ma) to establish ‘baseline’ conditions from which to explore
the effects of different climatic change scenarios resulting from the impact of
a large bolide. The post-impact scenarios contrast the effects of the possible
resulting short-term (global darkness, climatic cooling, high atmospheric CO™
concentrations) and long-term (high CO™, global warming) climatic change
that may have arisen owing to the impact.
Conclusions
The various global simulations described in this book provide representations of contrasting past and future climates and allow us to ask how the
physical and chemical nature of the Earth has influenced the functioning of
biological systems, with specific emphasis on terrestrial plant photosynthesis
7
8
Introduction
and the cycling of carbon through ecosystems. Some aspects of this work point
to the importance of climatic feedback on vegetation, its role in determining
other geochemical processes, and the possible responses of the terrestrial biosphere to abrupt environmental perturbations at the global scale.
We readily acknowledge that specific gaps exist in geological time which
remain to be investigated with a modelling approach and that some of these
gaps represent interesting intervals during the course of Earth history and
plant evolution. Among these, two in particular stand out: the late Silurian
and early Devonian, when the land became colonised by terrestrial plants; and
the end of the Permian (c. 249 Ma), recognised as the greatest mass extinction
in the history of life. The role played by large Devonian vascular land plants in
the cycling of carbon was probably substantial, with deep root penetration
into soils, and colonisation of drier upland and primary successional areas significantly enhancing rates of chemical weathering and atmospheric CO™
draw-down (Algeo & Scheckler, 1998; Elick et al., 1998).
In contrast to the early Devonian situation, the end-Permian mass extinctions would have severely altered the global carbon cycle through very different processes. Extinction of many groups of marine organisms would have
limited oceanic primary production whilst a major and sudden mass extinction of terrestrial floras, clearly evident in the high southern palaeolatitudes
(Retallack, 1995) and elsewhere, would have severely curtailed the terrestrial
carbon cycle. This ‘bottleneck’ in carbon cycling was probably further
increased by an extensive insect extinction at this time. As a result, decomposition was probably largely rate-limited by fungal activity, a feature evident in
the fossil record by the extraordinarily widespread occurrence of fungal
spores at the Permian–Triassic boundary (Visscher et al., 1996).
The development of more flexible and easier-to-use global climate models
should in the future allow model-based assessments of some of these intriguing issues. Preliminary palaeoclimate modelling studies for the late
Permian (Kutzbach & Ziegler, 1993), and general circulation climate modelling studies with different Palaeozoic continental configurations (e.g.
Crowley et al., 1993), offer promise for investigating how significant changes
in biological systems influenced biogeochemical cycles, particularly terrestrial
carbon cycle dynamics.